Many people now ask artificial intelligence to draw themselves by inputting their occupation, personality, preferences, and common speech patterns, generating images instantly, sometimes as fantasy characters or community post-style descriptions. Beneath the apparent joke lies sincerity: saying "Draw me" may really ask "How do you see me?" An AI-generated image is neither the real person nor solely another's gaze, but another self-portrait created by the machine combining provided words, data, and gaps between them. Results differ slightly each time even for the same person—one day appearing at a desk, another wandering a library—reflecting that humans can never be fixed as a single image. People have long used names, occupations, zodiac signs, blood types, MBTI, profile photos, and self-introductions to describe themselves; now prompts are added. The image AI draws may mirror the user but can never replace them; liking the crafted image too much might make one want to conform to it, yet people are always more complex than words and more off-kilter than images. The self-portrait in the prompt doesn't complete the user but reveals unfinished parts, prompting an age-old question: Who am I, and what do I want to look like?